There has been a striking resurgence of interest in surgery for Parkinson's disease (PD) with new targets identified and new procedures developed. This systematic review identified over 500 studies of surgery for PD published since 1990, including over 10 000 patients. However, the authors were unable to assess the value of PD surgery reliably because only seven randomised trials were identified including just 196 patients. Studies of surgery for PD have generally been of poor quality with too few patients, too short follow up, inappropriate choice of outcome measures, and lack of control groups. Much larger, randomised, controlled trials are needed to assess the longer term effects of surgery on patient rated quality of life and cost effectiveness.